The “Just One Thing” Series at Management Excellence is intended to provoke ideas and actions around topics relevant to our success and professional growth. Use them in good health and great performance!
—
There’s an interesting article in the May, 2014 issue of Harvard Business Review, entitled, “In Hiring, Algorithms Beat Instinct.”
According to the authors, we would be better served by letting algorithms do the heavy lifting before inserting our own bias-filled and easily distracted selves into the hiring equation.
The authors (Kuncel, Ones and Klieger) suggest, “…that a simple equation outperforms human decisions by at least 25%.” They offer that their findings extend to situations with large pools of candidates and at all levels from front-line to the C-Suite. They further cite our propensity to be thrown off by our cognitive biases, irrelevant data points, arbitrary comments in conversation and candidate compliments as reasons why we might need some objective help in our hiring decisions.
Provocative.
I’m experienced enough and comfortable enough in my own skin to recognize and agonize just a bit over my own hiring gaffes during the course of my career. Two in particular haunt my hiring dreams. In both cases, I would have taken a bullet for the decision on the front-end, only to discover in one case a fatal character flaw and in another, a fatal cultural incompatibility. An extra layer of insurance up front in the form of help from a reliable, predictive test instrument would have been highly valued in identifying and helping rule out these candidates before I made those costly mistakes.
However, (you had to know there was a “however” coming somewhere in this post) I truly struggle with the idea of deferring my up-front candidate pool crunching to an instrument. In particular, I fear missing out on the unique or outlier candidate that might be fairly wide of the algorithm’s parameters.
Many of my most successful hires have come from non-traditional backgrounds with very unique experiences to draw upon in our work. I’ve made a habit of avoiding HR screening of the candidate pool (particularly for strategic roles) in search of people with diverse and non-traditional backgrounds and success in analogous situations. While perhaps it’s a personal oddity of my own practices, I’m more interested in discovering remarkable people than I am in identifying perfect people.
The Bottom-Line for Now:
I’m intrigued by the help that these tools can provide. I’m also leery of suspending judgment and/or relegating my evaluation process to the candidates our screening instruments deem worthy of consideration. Some of the great contributors of my career wouldn’t have made it past HR without a hall pass from me, and I’m not willing or ready to relegate this to a program tuned by someone with a lot less interest, curiosity and drive to discover the next great contributor.
—
Don’t miss the next Leadership Caffeine-Newsletter! Register here
For more ideas on professional development-one sound bite at a time, check out: Leadership Caffeine-Ideas to Energize Your Professional Development.
New to leading or responsible for first time leaders on your team? Subscribe to Art’s New Leader’s e-News.
An ideal book for anyone starting out in leadership: Practical Lessons in Leadership by Art Petty and Rich Petro.
I agree completely. Often times it is that one candidate that doesn’t look good on paper, but after you talk to them you realize that there is a fire inside them that will make them outperform even the best looking candidate in the stack of resumes. I remember one particular developer that I hired, and on paper, there wasn’t that much there. You would remember the name if I dropped it. During our conversation, I discovered a lot more about why his work history looked the way it did, what motivated him, and what I could expect from him as a team member. Yes, there was some coaching to be done along the way and some rough edges to knock off (you should be starting to realize who I’m talking about), but this person has gone on to become a real technical leader and can outperform just about anyone in a software development role.
I think it’s worth the failures (and I’ve had my share too) in order to discover those great talents that would otherwise not pass the algorithmic filter. I believe in taking calculated risks, as those result in great leaps forward more often than not. Otherwise, go ahead, make the safe hires and condemn your organization to average growth and performance.
Dave, what a great add-on here. I’m not ready to let go of that opportunity…at least for the level and types of professionals we are both hiring. Well said! -Art
I would just like to point out that for every single exception you find, you will just as many misfires. Further, you are assuming human expertise would fail to pick up this passions whereas a well-designed test battery (which consists of structured interviews, biodata, and other standardized methods for assessing “passion”) would not. The point of the article is we can continue to stubbornly rely on human intuition, making sub-par hiring decisions, and then justify out actions by saying there was “this one time.” The point of research is that it uses many data points (e.g., statistics) rather than one (e.g., anecdote) to identify trends. In other words, the trends suggest you will hire more passionate, outstanding, and exceptional performers if you rely more on statistical algorithms than on human judgement.
Thanks Don for sharing your views! -Art
Art,
life or hiring or anything else, is about taking chances. When you take chances, you’re going to make mistakes. It comes with the territory. I suppose one could mitigate every risk and over analyze every option, but the cost incurred to prevent mistakes would so outweigh the gains made from chances taken, it would be silly.
There are many cases where people, usually people selling analytics tools, make these types of claims. But that comes from looking for difference after the results are already known, i.e. a rephrasing of the “survivorship bias”. You can’t just look at the people you hired and say if it was successful or unsuccessful, you’d also have to look at all the people you didn’t hire, and determine if they would have been successful or unsuccessful – and if you would have enjoyed working with them.
There are two types of statistics. Descriptive statistics, which explain what happened and Predictive statistics, which predict what will happen. These Harvard cats are taking descriptive statistics and saying they have predictive capabilities. They’ll take all the consulting and publishing dollars they can get selling that snake oil, but I bet they don’t sit down at many poker tables using it. If they do, I’d love to sit down at that table with them.
Andy
Great catch Art!
I agree with your assessment. As an executive recruiter, I see value in analytically assessment tools but not exclusively and less so at the front end. As you wisely state, too often these filters eliminate some great talent. In my experience, hiring success is directly driven from getting crystal clear on the target candidate profile (this is very serious business and most often glossed over as HR develops a Job Description). Then vetting that profile among the stakeholders in the hiring organization, making sure that the desired character traits are clearly defined and understood is also critical. As is with great leadership, great candidates and hires are born from the heart – the gut – the core of someone’s character and weather or not there is a basic cultural fit. You can’t rely on resumes and statistical tools to uncover that stuff – and if you do, you may be filtering out some really powerful talent. Solid quality of character can often trump specific experience.
The Harvard folks also miss a huge part of this. The source of hiring failure is not only found in the hire – If I had a nickle for every hiring organization that had its share of organizational, operational, and cultural dysfunction – well, I’d have a really good collection of nickels :). A good thorough hiring process will surface these dysfunctions and present them as the risks that they are so they can be mitigated. If the goal is a successful and productive hire, I find that this holistic approach yields the best track record of success.